COLORADO IRRIGATION WATERS AND THEIR CHANGES. 71 
the alkali incrustations, which appear in many localities; those of Long- 
Pond and the Windsor Reservoir resemble the water soluble portion of 
the soil, rather than the efflorescent alkalies. The water of Warren’s 
Lake has suffered less change in the character of the salts held in solu¬ 
tion than the others, and yet, the sulfates compose rather more than 50 
per cent, of the salts held in solution. 
19. The salts predominating in the water of the Poudre, while it is a 
mountain stream, are the carbonates, with some chlorids and sulfates, 
but as, stored in Terry Lake and Windsor Reservoir, the carbonates 
have almost disappeared and their place has been taken by the sulfates. 
20. The amounts of calcic magnesic and sodic sulfates which ap¬ 
pear in the stored waters are large. We find in Terry Lake 5,859 tons of 
calcic sulfate, 10,616 tons of magnesic sulfate and 7,1J3 tons of sodic sul¬ 
fate. In the Windsor Reservoir we have the same salts, but in different 
quantities, 6,083 tons calcic sulfate, 7,029 tons of magnesic sulfate, 1,999 
tons of sodic sulfate. These are the salts which constitute our alkalies^ 
21. The only constituent contained in these stored waters which, 
under our conditions, may have any great interest or significance as a 
plant food, and consequently tend to maintain the fertility of the soil, 
is the potash, K 2 0. The quantity of potash held by the stored waters 
is not great. The aggregate amount present in the four lakes discussed 
is 188 tons, contained in 27,672 acre-feet of water, which, allowing two 
acre feet of water to an acre of land, would give an application of fifty 
pounds of sulfate of potash to the acre, which undoubtedly tends to 
maintain the fertility of the land to which it is applied 
22. The potash contained in the stored waters is largely brought 
into the reservoirs by seepage or other than river water. 
23. The application of two acre-feet of river water as it flows through 
the canyon would give only 12.5 pounds of sulfate of potash per acre, or 
exactly one-fourth as much as the stored waters. As the seepage water 
contains not more than one-third of the latter in either of these cases, 
it follows that the amount of potash carried by it and neccessarily ob¬ 
tained from the soil through which it has seeped, is much greater than 
that carried by pure river water, and we may note that the quantity in¬ 
dicated is greater than that carried by drain water or by soil water, as a 
rule, but is less than that carried by off-flow water, and sometimes by 
soil water. 
24. The amount of nitrogen, including all forms, added with the irri¬ 
gation water, being less than four pounds per acre, is negligible. . 
25. The quantities of useless, or even deleterious salts,.added to the 
soil by the application of two acre-feet of stored water to an acre of 
land, are worthy of consideration. In the case of the Windsor Reservoir 
we add the equivalent of 54 pounds sulfate of potash, and at the 
same time 5,347 pounds of other salts; in the case of Terry Lake we add. 
55 pounds of sulfate of potash and 11,349 pounds of other salts. 
26. Water used in direct irrigation, that is, water conveyed by 
means of ditches directly from the river to the land irrigated, suffers 
less change than when stored, but does not by any means escape alto¬ 
gether. The best measure that I have of the extent of this change,, and 
one which, judging by the extent that the water supplied to Fort Col¬ 
lins is changed in flowing less than eight miles, is not an : extreme or an 
exaggerated one, indicates that the total solids are not less than five 
times as much as in the river water when the ditch was not more than 
ten miles long. 
27. The water used in irrigating, in order to study its changes, was 
water taken directly from the river, so far as we could obtain such. The 
general results may be stated as follows: 
28. The water flowing over the soil carries, in the first portions 
which flow off, very considerable amounts of salts in solution. The 
Samples which gave the most reasonable results indicated that water flow¬ 
ing for 6(H) feet over the plot experimented with, carried between *nn and 
1,000 pounds more salts in solution, per acre foot, than the on-flowing 
